While it is likely that having more catalogs, and in turn, having a
larger sample size, especially in the earlier decades, would have
yielded more robust trends and results for the earlier decades, it is
unlikely this would have changed the overall results.